Nobody funds year twelve without a year two

Two public funding instruments, two jurisdictions, one problem. Both exist because the return on a climate asset arrives long after the money goes in, and each solves it the same way, by putting something a third party can check into the years in between. That is also the fix for the copy, and it is missing from most of what gets written about these companies.

The Direction

On 16 October 2025 the Subsidy Advice Unit, part of the Competition and Markets Authority, published its report on the Contracts for Difference scheme the Department for Energy Security and Net Zero proposed for Allocation Round 7. A CfD does not pay a generator an income. Paragraph 1.11 describes a two-way mechanism: a top-up when the reference price sits below the strike price, and a payment back to the counterparty when it sits above. What the contract sells is a narrower range, which is the thing a lender prices.

The reform, at paragraph 1.16(c):

"Increasing the length of new CfD contracts from 15 to 20 years for fixed-bottom offshore wind, FLOW, remote island wind, onshore wind and solar technologies. This change helps rebalance costs between the short term (where major growth in investment is needed) and long term (where the benefits of that investment will continue to be experienced)."

Subsidy Advice Unit, 16 October 2025

Read that second sentence again as a sentence about your own company. The investment is needed now. The benefits are experienced later. Everything difficult about explaining a business like yours sits in the gap between those two clauses. At paragraph 2.27 the report records the purpose: the longer term "is intended to improve investor confidence". Five more years of price stabilisation, bought to make the early years financeable.

Then look at what the department says about the far end of its own instrument. At 2.57(a) the Assessment claims the extension will "exert a downward pressure on strike prices and decrease energy costs by approximately £1-2/MWh", and then, in the same sub-paragraph:

"However, DESNZ acknowledges that there are risks associated with the extended term such as long-term subsidy cost increases (post-2045) due to the uncertainty of future wholesale prices, and increasing potential competitive distortions"

At 2.38(b), on the same reform:

"An impact assessment referenced in the Assessments suggests the net cost implications of this are uncertain."

A pound or two per megawatt hour in the near term, with a figure on it. Beyond 2045, an acknowledged unknown, in a document whose entire purpose is to establish that the scheme is proportionate and limited to what is necessary. A department with statutory powers and the machinery of the state behind it can price the interval and cannot price the endpoint. If that is where a government lands, it is where your reader is standing too.

The second instrument is stricter, and it is the one that got tested hardest. In the Innovation Fund 2024 Hydrogen Auction, the second run under the European Hydrogen Bank, which "opened on 3 December 2024 and closed on 20 February 2025", projects competed for

"a fixed premium payment upon production of certified and verified hydrogen (EUR/kg RFNBO hydrogen), for a maximum period of 10 years of operation"

The Commission is explicit about when money moves, and about the clock that starts once a winner signs:

"Once their grant agreement is signed, projects must reach financial close within 2.5 years and start production within 5 years. Projects will receive payments for up to 10 years only upon certification and verification of renewable hydrogen production (i.e., not before Entry into Operation)."

Then watch what those conditions did to real projects. The auction "offered a total budget of €1.2 billion". On 20 May 2025 the Commission "selected and invited to grant agreement preparation 15 renewable hydrogen production projects requesting a total of €992 million in EU support". Winners then had to lodge a signed completion guarantee within two months, which the Commission describes as due diligence "to protect public funds from projects bids that are not sufficiently mature". What happened next is recorded on the same page:

"In light of this requirement, several of the 15 originally invited projects reassessed their submitted bids and maturity levels, and decided not to move forward with the signature of their grant agreements."

Projects were pulled up from the reserve list. By 20 January 2026, "six projects signed grant agreements under the IF24 Auction, receiving a total of €270.6 million in support". Nineteen of the projects invited into grant preparation are marked withdrawn.

The Explanation

That attrition is the most useful fact in this issue, and it is not an argument against the instrument. It is the instrument working. Roughly three quarters of the awarded value fell away at the point where a bid had to become a guarantee, and the Commission says in terms that the requirement is what prompted the reassessment. Everything that survived contact with the checkpoints is worth more than everything that did not, and now everybody can tell which is which.

Put the two side by side and the common structure is plain. A twenty-year contract carries a strike price and a delivery year. A premium arrives per certified kilogram, against a financial close at 2.5 years and production at 5. Both take a horizon no lender will underwrite whole and break it into intervals somebody can check.

Now look at how companies inside these regimes describe themselves. The pattern I keep finding is a claim about the endpoint and silence about everything before it. Gigatonnes by 2040. Cost parity at scale. A decarbonised process for an industry that has not changed since the 1960s. All of it may be true, and none of it is checkable this year, or next year, or in the year the reader is deciding.

What follows is my argument rather than a finding of any body quoted above, and I will not dress it as anything else. This is why long-horizon copy reads as faith. The size of the ambition has little to do with it. What is missing is a rung between here and the endpoint. A reader who cannot check anything in the interval has two options, which are to believe you or not, and belief is expensive to ask of somebody deploying capital.

The instruments show you what the missing rung looks like. In each case it is a fact with a date on it, produced by the company, verifiable by somebody else, that would look different if the thesis were wrong. Entry into operation, a megawatt in the ground, a tonne certified, or a cost per unit at a stated volume this year set beside the same number last year.

So the question to put to your own copy is narrow, and it is not whether the writing is compelling. It is this. What can somebody check in two years, and what would it look like if we were wrong?

I spent twenty five years as a barrister, advising clients through processes that ran longer than they wanted and cost more than they expected. You never hold a client in a case by telling them how it ends. You hold them by telling them what the next hearing will show, and what it would mean if it went the other way. A client who knows what to watch can wait. A client who has only been given the destination starts asking, around month nine, whether any of this was ever real.

Investors do the same thing, more politely, and usually to somebody else.

The Sentence

This one is real, and it is quoted because it is already right.

Verdagy's homepage leads on a status:

2 Megawatt Electrolyzer in Operation

and sets the destination directly underneath it:

Enabling zero carbon intensity, clean hydrogen at fossil parity by 2028.

Verdagy, read live 31 August 2026

A present-tense fact a third party could go and check, and then the horizon it is heading towards. That is this entire issue on one page, in the right order, and it cost them a line. A reader wondering whether 2028 is realistic has been handed the one fact that bears on it, which is a megawatt figure and the words "in operation".

The commoner shape, and the move that repairs it. Both halves of what follows are mine, written to demonstrate the move rather than lifted from anybody's page.

Our technology will remove carbon at under $100 a tonne at commercial scale.

A destination and nothing else, and the sentence every competitor in direct air capture is also writing. It identifies the category rather than the company.

The same claim with the interval put back in:

We are at $312 a tonne on our first commercial unit, down from $540 on the pilot, and the next unit is contracted for 2028. The curve gets us under $100 at the fourth build. If the third unit does not come in under $200, that timeline is wrong and we will say so.

Longer, and it does something the first cannot. It gives a reader a number to hold, a date to hold it against, and the condition under which the writer would concede the argument. The last sentence costs the most nerve and buys the most, because a claim that names its own falsifier is a claim somebody can act on.

You do not need permission to write this way. You need to be willing to publish a number your own next update might beat.

The Ruling

Two rulings from the same regulator, twenty-two months apart, and the distance between them is this issue in miniature.

On 7 June 2023 the ASA upheld a challenge it had brought itself against Petronas, over a television ad seen in September 2022 that closed on "achieve net zero carbon emissions by 2050", with "Net Zero 2050" on screen. The ruling opens by placing itself in context: it "forms part of a wider piece of work on environmental claims in the Energy sector, following intelligence gathered by the ASA".

Petronas answered the challenge in a sentence worth reading twice:

"They had taken steps to achieve net zero and would soon be announcing a pathway to progress and accelerate the aim."

The aspiration dated from 2020. The pathway was still forthcoming when the ad ran. Clearcast, defending the clearance, put it more plainly again: "The ad was about ambition and planned change."

The ASA set the destination against a present-tense number, recording that in 2021 Petronas's operations "produced 45.2 million tons of carbon dioxide and other greenhouse gases (GHGs)". Then the finding:

"We therefore considered information about the balance of Petronas' current activities, its emissions and the pathway to reducing them in line with the claims made in the ad, was material information likely to affect consumers' understanding of the ad's overall message and so should have been made clear."

The pathway to reducing them. A regulator treating the route between here and the target as material information rather than as optional colour.

Now the same regulator, twenty-two months later. On 9 April 2025 the ASA did not uphold complaints brought against a Shell television ad, including by Adfree Cities and Carbon Tracker. That ad carried superimposed text:

"In 2023, 68% of Shell's global investments included oil & gas, 23% included low-carbon energy solutions and 9% non-energy products. Shell's target is to become a net zero emissions (NZE) energy business by 2050"

A 2050 target with this year's percentages standing next to it. The ASA held that "the superimposed text enabled viewers to understand the relative balance of Shell's investments in higher and lower carbon activities at the time the ad appeared", and that the ad "had therefore not given a misleading overall impression of Shell's environmental impact".

Reading the two rulings against each other is mine, and the ASA did not draw the comparison. The same destination, the same decade, the same regulator. One company published the route and its present position alongside the target. The other was still going to announce the route. That difference cost about a line of type.

What can somebody check in two years? If you cannot answer that in one sentence with a figure in it, the problem is not that your horizon is too long. It is that you have not yet written the rung.

Next issue: carbon border pricing arrives, and your emissions number stops being a reason to like you and becomes a line in somebody else's cost model.

If you want help explaining something difficult, that is what I do.